Partitioning Simplification Study Of Casting Point Cloud Data Based On Curvature Features

With the development of industrial and computer vision technology,high-precision point cloud data obtained by 3D laser scanning technology can be used for the planning of cutting and processing trajectories of castings.However,the high-precision point cloud data obtained by 3D laser scanning has a large number of redundant points,which easily causes the waste of computer resources and brings impact on the speed of subsequent point cloud processing,so it is necessary to simplify the processing of the 3D point cloud data of castings.In this paper,for the industrial casting processing process,in order to improve computer efficiency,and enhance the retention rate of casting feature information,a curvature partitioning-based casting point cloud data simplification method is proposed,the main research work of this paper is as follows:(1)To address the problem that the large volume of casting point cloud data leads to slow calculation of geometric features,a method is proposed to construct the topology of point cloud data by dividing the original point cloud data into multiple sub-cubes with specified edge lengths through an octree encoding method,and then the curvature features of the point cloud data are estimated separately using the k-neighborhood method to ensure the efficiency of the point cloud simplification process.(2)In view of the large curvature of the working area of the castings,a simplification strategy of sub-region of the casting point cloud data by curvature features is proposed.According to the adjustable curvature threshold,the point cloud data is divided into multiple effective regions,and different simplification strategies are selected in different regions,combining the random sampling simplification method of preserving the boundary with the simplification method based on the regional center of gravity to simplify the casting point cloud data efficiently and accurately.(3)In the evaluation of the quality of point cloud simplification,the data quality of the proposed casting point cloud simplification method is evaluated from two aspects,namely,the visual comparison of point cloud simplification and 3D modeling results,by comparing this method with other three point cloud simplification methods.The experiments show that with the same simplification rate,the speed of the proposed method is increased by 29.9% and 33.8%,respectively,and the retention rate of feature points is increased by 15.1% and 19.2%,respectively,compared with the regional center of gravity method and the enclosing box method.After the verification of cutting experiments,the method in this paper can obtain accurate and effective simplified data of the casting point cloud,and the cutting error can meet the accuracy requirements of casting processing.